Machine Learning and 3D Image-Based Modeling for Real-Time Body Weight and Body Composition Estimation During Emergency Medical Care: Study 2

The performance of the 3D camera system needs to be evaluated in conditions in which it will be used in a clinical setting, but without the risk to patients. It also needs to be compared against other, standard, methods of weight estimation. In addition the downstream accuracy of drug doses, with the correct use of alternative dose scalars (such as ideal body weight or lean body weight) in patients with obesity needs to be evaluated. In this prospective crossover simulation study, resident and attending emergency physician volunteers will conduct simulated emergency care on simulated patient volunteers. The physicians will be randomised to start with either 3D camera systems, or standard care systems. After completing the scenario they will perform a new simulation scenario on a different patient using standard methods of weight estimation (crossover). At the end of the data collection session, each physician will have performed four scenarios (two standard methods, two 3D camera methods). Each simulated patient will similarly have participated in two standard method scenarios and two 3D camera scenarios. Each simulated patient and each physician will only have a single encounter. In this way the accuracy of weight estimation and drug dosing can be compared between the 3D camera system and standard methods. In addition, inter-rater reliability can be determined and compared between resident and attending physicians.